Prognostic-oriented fuel cell catalyst aging modeling and its application to health-monitoring and prognostics of a PEM fuel cell

Abstract

Today, poor long-term performance and durability combined with high production and maintenance costs remain the main obstacles for the commercialization of the polymer electrolyte membrane (PEM) fuel cells (PEMFCs). While on-line diagnosis and operating condition optimization play an important role in addressing the durability issue of the fuel cell, health-monitoring and prognosis (or PHM) techniques are of equally great significance in terms of scheduling condition-based maintenance (CBM) to minimize repair and maintenance costs, the associated operational disruptions, and also the risk of unscheduled downtime for the fuel cell systems. The two essential components of a PHM scheme for a general engineering system are 1) an accurate aging model that is capable of capturing the system's gradual health deterioration, and 2) an algorithm for damage estimation and prognostics. In this paper, a physics-based, prognostic-oriented fuel cell catalyst degradation model is developed to characterize the relationship between the operating conditions and the degradation rate of the electro-chemical surface area (ECSA). The model complexity is kept minimal for on-line prognostic purpose. An unscented Kalman filter (UKF) approach is then proposed for the purpose of damage tracking and remaining useful life prediction of a PEMFC.